Create README.md

README.md

@@ -0,0 +1,52 @@
+---
+license: apache-2.0
+tags:
+- genomics
+- dnabert
+- virology
+- foundation-model
+- hvilm
+---
+
+# HViLM-base: A Foundation Model for Viral Genomics
+
+This is the base pre-trained model for **HViLM**, as described in the paper:
+**"HViLM: A Foundation Model for Viral Genomics Enables Multi-Task Prediction of Pathogenicity, Transmissibility, and Host Tropism"**
+
+- **Paper:** [Link to your arXiv paper will go here]
+- **Fine-tuned Models:**
+    - `duttaprat/HViLM-finetuned-pathogenicity` (coming soon)
+    - `duttaprat/HViLM-finetuned-host-tropism` (coming soon)
+    - `duttaprat/HViLM-finetuned-transmissibility-R0` (coming soon)
+
+## Model Description
+
+(Paste your abstract here)
+
+## How to Use
+
+This model requires trusting remote code because it uses custom architecture files (`bert_layers.py`, etc.).
+
+```python
+from transformers import AutoTokenizer, AutoModel
+import torch
+
+repo_id = "duttaprat/HViLM-base"
+
+# This will download the files you just uploaded
+tokenizer = AutoTokenizer.from_pretrained(repo_id)
+model = AutoModel.from_pretrained(
+    repo_id, 
+    trust_remote_code=True  # <-- This is ESSENTIAL
+)
+
+print("Model and tokenizer loaded successfully!")
+
+# Example: Get embeddings for a sequence
+sequence = "ATGCGTACGT..." 
+inputs = tokenizer(sequence, return_tensors="pt")
+with torch.no_grad():
+    outputs = model(**inputs)
+    embeddings = outputs.last_hidden_state
+
+print(embeddings.shape)